Such a device with a bipolar transistor is frequently used as a switching transistor, for example in a static power conversion circuit. The switching speed of the transistor, which is usually a discrete component, should often be as high as possible.
Such a device and method are known from U.S. Pat. No. 4,969,027 published on Nov. 6, 1990. A bipolar transistor is disclosed therein (see FIG. 1B) which is provided with means for limiting the saturation. Said means comprise a Schottky clamping diode integrated into the device and formed in that the connection conductor of the base region is also put into contact with the collector region. Since the Schottky diode limits the degree of saturation, the switching time of the transistor is as short as possible, while on the other hand the latter does not get very far removed from the saturated state, which implies that a high collector-emitter voltage and high losses in the on-state are avoided.
A disadvantage of the known device is that it is less suitable for mass manufacture. Such a manufacture renders it important that the properties of the device within a batch and among batches do not vary too widely. The leakage current of a Schottky diode, however, is often too variable, in particular at higher temperatures. This is because the surface of the semiconductor body has a major influence on said leakage current, and because the properties of a surface are often difficult to control. A further influence here is the occurrence of so-called spikes, i.e. locations where the metal of the connection conductor of the base region penetrates into the collector region.
It is an object of the present invention to provide a device which does not have the above disadvantage and which comprises a bipolar transistor which is provided with means for reducing the degree of saturation, which is suitable for mass manufacture, and which has excellent properties as regards the leakage current.
According to the invention, a device of the kind mentioned in the opening paragraph is for this purpose characterized in that the second connection conductor is exclusively connected to the second semiconductor region for the purpose of preventing saturation of the transistor, and in that a partial region of that portion of the second semiconductor region which lies outside the third semiconductor region, as seen in projection, and which lies adjacent the second connection conductor comprises a smaller flux of dopant atoms. The invention is based on the recognition that the region where the doping flux has been reduced in such a device acts as a base diode which is connected in parallel to the collector-base junction of the transistor. When the transistor is in the saturated state—i.e. both the base-emitter and the base-collector junction are forward biased—the saturation collector current is mainly defined by the base-emitter voltage, and the saturation base current by the base-collector voltage. Since the base diode also passes part of the current through the base connection conductor, the base current is reduced. This reduces the base-collector voltage, with the result that the quantity of charge stored in the collector, i.e. in the weakly doped portion of the collector, is reduced. This reduces the switching time of the transistor. A further result of the above is that the saturation collector-emitter voltage rises because the collector-base voltage becomes less negative. An exchange takes place as it were between switching time and saturation collector-emitter voltage. The degree of saturation in a device according to the invention is limited by the base diode. The operation of the partial region having a reduced doping flux may be explained as follows: the current through the base diode is proportional to the gradient in the electron concentration of the base connection to the base-collector junction of the base diode. This gradient can be increased in that the thickness of the relevant partial region of the base is made smaller (the denominator of the gradient becomes smaller), or in that the doping concentration (p) is made smaller, so that—given a constant pn product—the electron concentration (n) at the junction becomes higher (the numerator of the gradient becomes greater), or by a combination of both measures. It is assumed in the above that the electron concentration adjacent the second connection conductor is zero, or at least very small, because of the very high recombination rate prevailing there. In all cases, accordingly, the desired behaviour of the device according to the invention is realized through the formation of a base diode which arises owing to a local reduction in the flux of dopant atoms in the base region of the transistor. This base diode may be regraded as a pn clamping diode. It does not have the disadvantages of a Schottky clamping diode mentioned above as used in the known device. As a result, a device according to the invention has a narrow spread and a good reproducibility as regards its leakage current. The characteristic formulated above, moreover, implies that the properties of the partial region of the second semiconductor region should be formed separately. This renders it possible to adapt the properties of the pn clamping diode, if so desired. This again leads to the possibility inter alia of obtaining a combination of the switching time ts and another important transistor characteristic, i.e. VCEsat (the collector-emitter voltage in the on-state) suitable for a given application.
In a first embodiment of a device according to the invention, the local reduction in the flux of dopant atoms is realized in that the doping concentration of the second semiconductor region is locally reduced. In a second embodiment, this is realized by means of a localized small thickness of the second semiconductor region. Preferably, both measures are taken at the same time. The flux of dopant atoms in the partial region of the second semiconductor region should be great enough in all cases for preventing punch-through. This is to say, it should be great enough to prevent the depletion layer of the base diode from reaching the second connection conductor at the maximum admissible reverse voltage. The leakage current of the pn clamping diode could otherwise become variable and high in that case, as in a Schottky clamping diode. In practice, very wide possibilities remain for making both the thickness and the doping concentration of the partial region substantially smaller than in the remainder of the portion of the second semiconductor region which lies outside the third semiconductor region. It is necessary for a good operation of the pn clamping diode that the latter should be in the vicinity of, and preferably below the second connection conductor.
In another very favourable embodiment of a device according to the invention, a fourth semiconductor region of the first conductivity type is present between the partial region of the second semiconductor region and the second connection conductor, which fourth semiconductor region has a small thickness and a high doping concentration, preferably the same thickness and doping concentration as the third semiconductor region. The device is then as it were provided with a special version of a clamping transistor. Its effectivity as a means of limiting the degree of saturation then also depends on the remaining portion of the partial region of the second semiconductor region which has a reduced flux of dopant atoms. The adjustment of the thickness of the semiconductor region thus provides an additional degree of freedom for the adjustment of the effectivity of the saturation limitation. The doping concentration of the fourth semiconductor region has substantially no influence on said effectivity. If the fourth semiconductor region has the same thickness and doping concentration as the third semiconductor region, the important additional advantage is obtained that the fourth semiconductor region can be provided simultaneously with the third semiconductor region, which keeps the manufacture simple. The presence of the fourth semiconductor region also provides a further protection for the device, in particular the clamping device, against the influence of the spikes mentioned earlier.
A method of manufacturing a semiconductor device whereby a first semiconductor region is formed in a semiconductor body comprising a substrate, which first semiconductor region lies in the semiconductor body, is of a first conductivity type, forms a collector region of a bipolar transistor, and is provided with a first connection conductor, whereby a second semiconductor region of a second conductivity type opposed to the first is formed above said first semiconductor region, which second semiconductor region forms a base region of the transistor, adjoins the surface of the semiconductor body, and is provided with a second connection conductor at said surface, whereby a third semiconductor region is formed which is recessed into the second semiconductor region, which is of the first conductivity type, which forms an emitter region of the transistor, and which is provided with a third connection conductor, and whereby the device is provided with means for preventing a saturation of the transistor during normal use, is characterized according to the invention in that the second connection conductor is exclusively connected to the second semiconductor region for the purpose of preventing a saturation of the transistor, and in that a partial region of that portion of the second semiconductor region which lies outside the third semiconductor region, as seen in projection, and adjacent the second connection conductor is provided with a smaller flux of dopant atoms. A device according to the invention is obtained by such a method.
In a preferred embodiment, the partial region of the second semiconductor region is formed below the second connection conductor and is given a smaller thickness and a lower doping concentration. Preferably, the partial region of the second semiconductor region is formed by means of ion implantation. This renders it possible to adjust the properties of the partial region in an accurate and reproducible manner.
In a favourable modification, a thin fourth semiconductor region of the first conductivity type is formed between the partial region of the second semiconductor region and the second connection conductor, preferably simultaneously with the third semiconductor region. A device according to the invention is thus obtained in a simple manner, comprising a special modification of a clamping transistor as described above.
The invention will now be explained in more detail with reference to two embodiments and the accompanying drawing, in which